AI Chips news and analysis on Web Pulse: 6231 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
Nvidia's Blackwell architecture achieves 20 times more AI agents per megawatt than its Hopper generation, according to the new AgentPerf benchmark. The efficiency leap, driven by FP4 precision and advanced NVLink designs…
NVIDIA unveiled the RTX Spark superchip for Windows PCs at GTC Taipei and Computex, delivering 1 petaflop of AI performance and up to 128GB of unified memory. The chip, developed with MediaTek, will power slim laptops an…
NVIDIA achieved leading agentic coding performance on the first agentic AI benchmark, AA-AgentPerf, delivering up to 20x better performance than previous generations. The benchmark, created by Artificial Analysis, measur…
Marvell Technology hired Adobe CFO Dan Durn, signaling a shift in investor focus from software to semiconductor companies in the AI era. Adobe shares fell 5% despite raising revenue guidance and tripling AI revenue, as t…
Nvidia CEO Jensen Huang announced a paradigm shift in computing from data retrieval to real-time generation, driven by AI factories and purpose-built CPUs. The company's Vera CPU for agentic AI has already generated $20 …
SK Hynix plans to triple memory production, but SK Group chair warns the expansion will take up to eight years, highlighting the slow pace of semiconductor manufacturing scale-up amid surging AI demand.
NVIDIA's Blackwell Ultra NVL72 platform achieved top performance on the first agentic AI benchmark, AgentPerf from Artificial Analysis, running up to 20 times more agents per megawatt than the previous Hopper architectur…
The AI industry is increasingly entangled in a web of circular deals where companies like Nvidia, Microsoft, and Anthropic invest in each other while also buying and selling AI chips and cloud services. This interdepende…
AXT Inc. shares surged 10% after Chinese export permit delays tightened global supply of indium phosphide, a critical substrate for AI data center optical interconnects. The company, which controls nearly 80% of the worl…
Four RTX Pro 6000 Blackwell GPUs in a single water-cooled rig failed during sustained model training when one card repeatedly dropped off the PCIe bus under load. The cause was a missing power inductor that came off the …
AMD opened pre-orders for its Ryzen AI Halo developer platform, a mini PC powered by the Ryzen AI Max+ "Strix Halo" processor with official Linux support and up to 128GB of unified memory for running large AI models. The…
Apple plans to launch four new Mac models later this year, including a Mac Studio with M5 Max and M5 Ultra chips, an M5 iMac with a refreshed color palette, and a Mac mini with M5 and M5 Pro options. The company is also …
China opened its first dedicated photonic computing laboratory in Shanghai, a joint venture between Shanghai Jiao Tong University and startup Lightelligence, signaling Beijing's bet on light-based chips as a strategic wo…
SpaceX rented its Colossus 1 data centre to Anthropic after latency issues, aging network infrastructure, and a mix of Nvidia chip generations prevented the facility from effectively training its Grok AI models. The Memp…
London Tech Week 2026 concluded its 12th edition on Friday, drawing over 30,000 attendees from 130 countries to Olympia and fringe events across the city. The UK government announced a £1.1bn AI Hardware Plan and a £200m…
UCLA Samueli School of Engineering launched a $125 million semiconductor hub with Broadcom, Applied Materials, GlobalFoundries, Meta, and Synopsys to accelerate AI-powered chip research and workforce development. Dean Al…
South Korean President Lee Jae Myung said Friday in Rome that South Korea and Italy are "optimal partners" for business cooperation, capable of building a "new industrial order" together. Lee spoke at a business roundtab…
Qualcomm is reportedly deepening ties with Amazon Web Services around its AI200 accelerator, which supports up to 768GB of memory per chip and is slated for a 2026 rollout, according to a Wells Fargo research note cited …
Elon Musk's xAI has leased its entire 220,000-GPU Colossus 1 supercomputer in Memphis to rival AI firm Anthropic for roughly $1.25 billion per month, after the facility proved unsuitable for training its Grok models due …
Oracle Corp. reported a $54.75 billion capital expenditure surge in fiscal 2026 as it builds five massive datacenters totaling 5.5 gigawatts of capacity to support AI processing, with a $638 billion revenue backlog tied …